Computational Science and Its Applications – ICCSA 2026: 26th International Conference, Braga, Portugal, June 30–July 3, 2026, Proceedings, Part II(16768 Lecture Notes in Computer Science)

The 3-volume set, LNCS 16767, 16768, 16769, constitutes the refereed proceedings of the 26th International Conference on Computational Science and Its Applications – ICCSA 2026, held in Braga, Portugal, during June 30–July 3, 2026.

The 116 full papers, 3 short papers, and 2 PHD showcase paper were carefully reviewed and selected from 405 submissions. The papers have been organized in topical sections as follows:

Part 1 : Computational Methods, Algorithms and Scientific Applications; High Performance Computing and
Networks; Geometric Modeling, Graphics and Visualization; Advanced and Emerging Applications.
Part 2 : Information Systems and Technologies; 
Part 3 : Information Systems and Technologies; Urban and Regional Planning; PHD Showcase Papers; Short Papers.

.- Information Systems and Technologies.
.- LLM-based Applications for Healthcare: A Systematic Mapping Study.
.- GazeMIL: Attention-Based Multiple Instance Learning for Interpretable Dyslexia Detection from Multi-Task Eye-Tracking Data.
.- Soft Routing-Inspired Specialization for E cient Satellite Image Time Series Segmentation.
.- Ending the War Between Model Selection and Combination: A Cross-validation Approach.
.- Misbehavior Detection in VANETs using Tree-Based Machine Learning Techniques.
.- Systematic Review of PPCP Models, Arti cial Intelligence, and Optimization Applied to the Small-Scale Textile Industry.
.- DepthGate: Condence-Gated Depth Veri cation for Multimodal Industrial Anomaly Detection.
.- Impact of Anatomical Positioning Markers on Breast Cancer Detection Thermography.
.- Predicting Temporal Metrics in Dynamic Systems Using Machine Learning.
.- Implementation of an Orchestrated Multi-Agent System for PPCP in Textile MSMEs: A Prompt Engineering Approach.
.- Adversarial Detection in EEG-Based BCIs: A Comparative Study of Classical and Neuro-Fuzzy Approaches.
.- DeOP-Light: Analyzing the Fine-tuning Paradox in CLIP Visual Encoder Compression for Open-Vocabulary Semantic Segmentation.
.- A Model Implementation to detect Hate Speech associated with Gender Violence by Natural Language Processing and Machine Learning.
.- Behavioral nance in determining the nancial choices of households.
.- A Multidimensional Quality Assessment of Synthetic Movement Data for Post-Stroke Rehabilitation.
.- SingLoRA-CLIP: Highly e cient VLM Adaptation with a Single Parameter Matrix.
.- Robust Vehicle Detection Framework Using YOLO12 with Advanced Augmentations for Adverse Weather Conditions.
.- A Multi-Objective Loss Function for Neighbor-Based Recommender Systems.
.- Algebraic Sheaf Convolution Networks for Local-to-Global Semantic Consistency in Heterogeneous Graphs.
.- Horizon-Aware Feature Selection and Evolutionary Optimization for Intraday Prediction in Futures Markets.
.- C-MTSS: Compact Multi-Teacher Single-Student Framework for Real-Time Multimodal Industrial Anomaly Detection.
.- Class-Speci c Feature Weighting Query Strategy for Online Active Learning in Intrusion Detection.
.- A Formal Constraint-Based Method for Detecting Gaps in Input Validation Test Suites.
.- MuSEvo: Assessing the Robustness of Multilingual Summarization Metrics through Evolutionary Adversarial Natural Language Processing-based Attacks.
.- Bayesian Analysis for Wild re Classi cation in the United States and Australia through Spatial Hierarchical Models.
.- ESGRep: Learning Domain-Speci c Embeddings for ESG Retrieval-Augmented Question Answering.
.- An Intelligent eUPF for Time-Sensitive Path Selection in B5G Edge Networks.
.- GPS: GlobalCLIP-PatchCore-SAM Based Zero-Shot Anomaly Detection and Localization in Smart Manufacturing.
.- Neural Networks and Gaussian Regression Process Comparison on a Physical Problem.
.- A Generalized AI-Powered Framework for CrossDomain Technology Readiness Assessment.
.- A Compact QSAR Model for BRAF Inhibitors Based on Gaussian Process Regression.
.- Machine LearningGuided Stochastic Switching Modeling of Chronic Kidney Disease.
.- A LightGBM Framework for Operational Tide Forecasting and Predictive Imputation in the Río de la Plata.
.- Enhancing Model Generalization in Index Futures Markets via Evolutionary Optimization and Data Leakage-Free Feature Engineering.
.- Zero-Shot LLM Sentiment Meets Market Microstructure: Parsimonious Feature Selection for High-Frequency Bitcoin Forecastings.
.- Evaluating Segmentation and Temporal Resolution in Highway Crash Forecasting: The BR-040 Corridor.
.- Semantic Noise Filtering in Post-Stroke Kinematic Data: An Automated Deep Learning Approach.
.- Semantic Leakage in CKKS via Homomorphic Noise.